►
From YouTube: Keptn Community & Developer Meeting - Month Day, 2021
Description
Regular sync-up in the "Entity Custom Resource Definitions for Keptn GitOps" project being done as a part of Google Summer of Code by Eric Y. Kim
Project page: https://github.com/keptn/community/blob/main/mentorship/gsoc/2022/projects/gitops-entity-CRDs
A
Okay,
so
hello,
everyone
thanks
so
much
for
joining
me
today,
for
I
guess
the
first
bi-weekly
meeting
for
the
entity.
Custom
resource
definitions,
google,
summer
code
project,
so
since
this
is
the
first
meeting
I'll
quickly
go
over
sort
of
what
I
hope
to
achieve
and
the
purpose
of
such
meetings.
A
So
each
of
these
meetings
are
designed
to
be
about
30
minutes
long,
and
it's
really
a
way
for
me
to
provide
an
update
either.
You
know
live
or
asynchronously
via
recording
on
sort
of
the
prog
on
on
my
progress
with
the
project
so
far,
as
well
as
the
space
for
discussion
on
on
various
things
with
regards
to
the
project,
and
I
think
it'll
also
be
like
a
good
way
to
document
my
work
as
as
the
summer
progresses.
A
So
that's
sort
of
the
purpose
of
these
meetings.
As
I
mentioned
today's
the
first
one
june
14th,
and
this
I'm
calling
it
the
coding
period,
kickoff.
A
Which
is
sort
of
yeah
the
exciting
part
of
of
the
whole
project,
where
I'll
begin,
the
implementation
of
the
project
itself.
A
So
today,
there's
sort
of
three
main
things
I
want
to
go
over,
the
first
being
just
a
quick
overview
of
or
just
a
quick
review
of,
the
community
bonding
period,
which
was
the
last
few
weeks
as
well
as
some
of
the
administrative
things.
This
part
might
be
a
bit
boring,
but
just
some
documentation
as
well
as.
A
Some
other
places
where
either
mentors
or
everyone
else
can
stay
updated
with
my
progress
and
and
sort
of
see
the
actual
the
code
itself
and
then
finally,
what
I'll
be
doing
the
next
couple
weeks
before
the
next
meeting,
which
is
the
implementation
of
the
service
level,
indicator
indicators,
operator
and
custom
resource
definition
and
actually
do
have
some
questions
about
this
as
well,
that
that
perhaps
we
can
discuss
either
either
in
this
meeting
or
or
on
on
slack
as
well.
A
Okay
and,
of
course,
feel
free
to
like
stop
me
or,
if
there's
anything,
any
questions
or
any
other
things
sort
of.
We
can
discuss
as
we
go.
A
So
the
first
thing
is
with
regards
to
the
community
bonding
period.
So
I
spent
a
lot
of
my
time
in
the
last
couple
weeks
setting
up
a
lot
of
the
meetings
big
shout
out
to
suraj
for
help
with
with
setting
those
up
so
setting
up
meetings
such
as
this
one
like
means
of
communication,
as
well
as
documenting
progress
which
I'll
go
over
in
the
administrative
section
as
well
as
I
spent
some
time
learning
the
opera.
A
The
operator
sdk
framework,
which
is
what
the
project
is
built
using
and,
of
course,
cube
builder,
which
is
sort
of
the
underneath,
the
hood
of
operator
sdk.
What
what
it
uses
and
also,
I
spent
some
time
elaborating
on
the
sli
crds,
our
crd
and
operator.
I
guess
this
is
the
controller
to
get
which
is
kind
of
in
progress.
A
I
will
go
over
kind
of
it
right
now,
but
I
do
have
some
questions
on
the
boat
that
later
that,
I
think
maybe
will
help
me
sort
of
actually
really
hash
this
out,
but
that's
that's
kind
of
how
I've
been
spending
my
time
as
a
quick
review
like
it's
a
little
bit,
I'm
a
little
bit
behind
in
terms
of
what
I
wanted
to
achieve
in
the
community
bonding
period,
but
I
also
in
my
schedule,
put
in
a
weak
buffer
just
in
case
there
were
some
other
things
I
couldn't
achieve
in
that
time.
A
So
if
you'll
notice
in
in
sort
of
my
timeline,
the
very
first
task,
which
is
this
is
actually
three
weeks
instead
of
two
weeks,
and
that
was
just
in
case
something
came
up
and
or
if
I
needed
a
little
bit
more
time,
for
example,
to
elaborate
on
on
the
ticket
itself
yeah.
So
that's
a
quick
update.
Is
there
any
comments
or
questions
about
that
just
feel
free
to
chime
in
or
put
in
the
chat?
A
If
so,
if
not
I'll,
move
to
sort
of
the
next
section,
which
is
just
an
overview
of
where
a
lot
of
the
documentation
lives
so
that
you
know
even
outside
of
of
these
two
week,
meetings
as
well,
as
I
mean
obviously
you're,
always
you're,
always
free
to
like
ping
me
on
slack
or
or
comment
on
on
any
of
like
the
github
issues,
but
just
in
case
I
just
just
for
some
something
asynchronous
updates.
A
I
did
create
some
documentation
which
I'll
go
over
really
quickly
and
in
fact
you
can
find
them
all
pinned
on
the
kept
and
operator
kept
in
github's
operator
working
group
channel
specifically
on.
I
guess
here
so
a
lot
of
so
basically,
everything
kind
of
the
main
sort
of
documentation
is.
Is
this
link
right
here,
which
is
the
main
project
page
that
lives
in
the
community
repository?
A
A
So
this
is
actually
a
wiki
on
the
captain
get
offs
operator
repo
which
has
sort
of
my
daily
progress
just
so
that
I
can
keep
just
so
that
just
for
my
own
documentation
sake
as
well
as
to
keep
everyone
updated
on
just
just
as
a
way
to
like
quickly
reference
what
I've
been
working
on
on
sort
of
a
daily
basis.
A
So
I
started
that
on
june
12th,
which
was
the
beginning
of
the
the
coding
period
so
you'll
see
like
each
of
the
days
we'll
have.
What
did
I
do
yesterday?
What
will
I
do
today
and
what
are
my
blockers,
and
this
will
I'll
do
this
basically
every
day
until
the
end
of
or
as
as
almost
every
day
until
the
end
of,
I
guess
the
conclusion
of
g-stock.
A
B
A
B
A
Yeah
yeah,
I
think
yeah,
that's
that's
true
too.
I
I
kind
of
like
the
wiki
format,
because
of
exactly
your
point
where,
like
you
can
because
I
think
this
will
get
very
very
big
like
it'll,
be
yeah,
so
maybe
especially
yeah.
I
think
this
might
be
a
better
means
of
documentation,
but
I
think,
like
that's
so
like
what
I've
been
basically
doing.
So,
for
example,
I
had
a
blocker
here
I
just
like
with
pain
old,
like
on
slack
and
like
he
would
help
me.
I
think,
that's
okay
for
now,
yeah
I'll.
A
Awesome
yeah,
so
that's
that's
kind
of
the
the
stand-ups
themselves.
I
don't
know
if
this
will
be
super
useful
for
the
mentors,
but.
B
I
think
I
think
it
will
be
you
you
kind
of
see
what
we'll
be
able
to
get
at
into
what
you're
doing.
Obviously,
I
think,
I'm
not
sure
if
the
mentors
are
going
to
check
this
every
day,
but
I
think
this
would
be
a
really
good
thing
if
having
this
publicly
is
a
good
thing
in
my
opinion,
because
when
you
are
like
applying
for
charts
and
stuff,
then
you
know
the
effort
kind
of
shows
up
in
all
of
this.
A
A
Okay,
sweet
yeah,
that's
good
to
hear
so.
Yeah,
definitely
we'll
be
updating
this
on
the
daily
or
as
much
as
I
can
on
the
daily.
A
For
sure
for
sure,
I
think
it's
just
a
good
way
for
me
to
like
stay
accountable
as
well,
just
with
like
everything's,
asynchronous
and
and
sort
of
remote
and
even
in
different
time
zones.
So
I
thought
this
was
a
good
way
to
make
sure
that
I
keep
myself
accountable,
also
on
top
of
like
the
documentation
and
keeping
everyone
updated.
A
Awesome
yeah,
so
that
is
the
daily
stand-ups
there's
also
here
a
link
to
the
calendar.
Of
course,
this
has
all
of
the
meetings,
including
our
stand-ups,
a
lot
of
other
g-stock
meetings
as
well,
which
is
awesome,
and
then
the
weekly
meeting
minutes.
A
So
once
again,
this
this
page
sort
of
mirrors
a
lot
of
what's
on
on
this
message
right
here,
but
but
I
would
say,
like
this-
would
be
the
most
sort
of
up-to-date
place
to
go.
A
But
I
don't-
I
don't
foresee
much
change
here,
other
than
like
recordings
being
added,
so
obviously
the
meeting
minutes,
which
is
the
document
that
we're
currently
on
as
well
as
I
guess,
what
I
want
to
segue
next
into
is,
is
sort
of
the
project
board
which
has
like
a
kanban
board
of
of
issues
that
I'll
be
working
on
or
that
I'll
be
like
elaborating
on
further.
A
A
So
I've
created
a
github
issue
for
that
here
and,
as
you
can
see,
it's
it's
in
terms
of
the
actual
api
definition
and
the
controller
itself.
It's
actually
not
super
complicated,
mostly
because
one
the
slide,
the
the
captain
slis
the
kept
in
sli
configuration
it
really
composed,
is
really
composed
of
of
two
things,
which
is
the
configuration
for
the
provider
itself,
as
well
as
the
indicators
which
are
just
like
the
queries
for
each
for
each
indicator.
A
In
here,
as
well
as
in
there,
so
my
controller
will
go
will
be
added
to
the
controllers
here,
as
well
as
the
definition
type
or
the
the
yeah.
The
entity
type
will
will
go
in
here,
which
is
actually
really
quite
easy
to
do
via
operator
sdk,
so
obviously
I'll
be
leveraging
that
a
lot
yeah,
so
so
the
ticket,
I
guess,
is
part
of
the
project
board
as
well
as
is
is,
is
an
issue
on
on
the
project
repo.
A
Some
things
I
still
need
to
finish
elaborating
on
is
with
regards
to
the
sli
provider,
I'm
a
little
bit
because
I
know
there's
currently
or
from
the
documentation.
It
appears
that
there's
currently,
two
providers
being
supported
in
addition
to
you
can
create
custom,
but
you
can
also
have
like
custom
sli
providers
for
for
the
sls,
so
for
the
api
definition
itself,
I'm
I'm
a
little
bit.
A
B
I
I
had
a
question
around
this
actually
so
yeah
when
we
we
have.
Why
do
you?
Why
do
you
say
that
the
sli
spec
is
dependent
on
a
slight
provider?
A
So
from
I
guess,
my
understanding
of
like
an
sli
or,
like
my
understanding
of
I
guess,
the
indicators
of
the
sli
is
that
they're
specific
to
the
the
quarries
themselves
are
specific
to
the
provider,
so
they're
actually
like
not
necessarily
defined
by
kept
in,
but
by,
for
example,
like
dyna,
trace
or
or
prometheus.
B
Okay,
sorry
yeah.
As
far
as
I
know,
the
sli
conflict,
that
is
the
sli
spec,
is
sort
of
independent.
What
changes
is
what
goes
inside
the
indicator
and
I'm
just
trying
to
find
something
that
would
confirm
our
assertion.
B
That
is
my
assertion
that
that
is
how
it
is,
but
I
I
think
we
do
not
have
a
dependency
on
provider
in
the
sense
that
when
you
create
an
sli
resource,
the
provider
specific
stock
goes
into
indicators,
which
is
a
list
of
key
value,
pays
like
you've
correctly
written,
but
I
don't
think
we
have
a
provider
specific
configuration.
B
Yeah,
so
the
query
key
so,
depending
on
what
you're,
using
for
example,
promise
yes
has
a
different
way
of
writing
the
query
right.
So
the
different
way
data
dock
is
a
different
way.
Similar
logic
is
a
different
way
right.
The
change
goes
into
indicators,
which
is
a
key
and
a
value
which
is
string
so
that
the
string
is
what
keeps
on
changing,
but
everything
is
there's,
there's
nothing
specific
to
provider
as
well.
B
I
think
the
the
reason
I'm
saying
this
is
because
that
would
vastly
simplify
our
implementation,
because
if
you
go
in,
if,
if
it
turns
out
that
it's
dependent
on
sli
provider,
then
you'll
have
to
handle
cases
for
different
providers,
and
I
don't
think
the
sli
context
was
designed
like
that.
A
Yeah,
I
think
for
for,
like
the
sl
so
like
I
guess,
when
you're
like
adding
an
sli
to
to
captain
it's
like
part
of
it
is
you
have
to
specify,
like
the
actual
configuration
of
the
provider.
A
B
A
Yeah,
no
sorry
I
was
just
saying
like
yeah,
I
think
what
I
was.
I
think
what
I
what
I
mean
by.
A
This,
so
maybe
I
should
rename
this
to
you
is.
I
was
kind
of
thinking
about
like
yeah
reason,
there's
like
a
resource
so
like
a
resource,
sli,
config,
yaml
and
that's
different
from
like
the
actual
sli
configuration
with
the
indicators.
I
think
so
like
I
think,
and
I
thought
and-
and
I
was
yeah
like
I
guess
my
impression
was
that
this
was
basically
whatever
the
provider
configurations
like.
I
think
some
of
this
is
just
like
environment
variables,.
B
Yeah,
that's
that's
what
I
think
dude.
So
I
think
the
sli
is
independent
of
everything
I'm
going
to
dig
in
through
the
captain
code
and
find
us
a
hard
proof
for
that.
Okay,
we
can.
A
A
I
almost
think
that
it's
it's
kind
of
a
not
I
guess
like
I
guess,
because
kept
in
itself
will
not
won't
handle
the
the
status
or
the
health
of
the
sli
providers
themselves.
So
then
I
don't
see
any
reason
why
the
controller
should
should
reconcile
that.
A
So
in
that
case,
this
might
actually
not
do
do
much,
but
I
think
either
way
it
should
like.
There's
the
the
I
guess
upon
like
initialization.
It
should
check
for
the
sla
configs
in
the
manifest
anyways
for.
A
Operator
yeah,
so
that's
sort
of
what
I
have
for
this
ticket.
I
I
don't
know
if
it
will
be
again
like
I
gave
myself
an
extra
week
for
the
very
first
task.
So
technically
I
guess
in
my
initial
timeline.
I
have
three
weeks
to
finish
this,
but
I
don't
think
I'll
take
that
long.
So
my
hope
is
that
by
the
next
meeting
I'll
be
able
to
to
sort
of
demo
the
the
crd
and
an
operator
which
yeah,
I
guess
in
a
couple
weeks
yeah
I
did
have
some
questions.
A
I
guess
this
one
we
we
just
dealt
with,
which
is
like
the
difference,
is,
I
guess
it's
not
con.
Well
yeah,
like
configuration
details,
yeah.
B
I
I'm
pretty,
I'm
pretty
sure,
there's
a
lie
is
independent
of
the
provider
and
the
reason
the
reason
I
say
this
is
because
a
dyna
trace
from
it
is
and
any
other
custom
integration
is
going
to
be
implemented
by
different
than
different
people.
So
if,
if
you're
using
the
same
sli
config
and
if
it
has
something
that
is
specific
to
one
integration,
for
example,
let's
say:
there's
a
field
specific
to
dynatrees.
That
means
prometheus
service
will
also
will
be
using
the
same
struct.
B
Essentially
because
it's
sli
and
then
it
would
have
to
care
about
what
to
do
with
that
field,
which
doesn't
sound
like
good
design
to
me.
So
I'm
pretty
sure
we're
not
using
any
things
like
anything
which
is
specific
to
the
provider
goes
inside
the
indicator,
which
is
the
query
specific
to
the
provider
and
that's
what
you
saw
in
sli
dot
yam.
So
we
don't
have
to
get
it
from
the
provider.
A
B
I
I
don't.
I
don't
think
that
would
happen
right,
because
sli
and
slo
are
like
independent
things.
As
in
they're
not
tied
to
a
provider,
the
the
construct
themselves
are
independent,
it
indicates
service
level,
a
level
indicator
and
service
level
objective
right,
so
that
should
be
independent
of
the
whatever.
Whatever
sli
provider
you
use.
Okay,
perfect,
so
I
don't
think
that's
going
to
happen
in
the
future,
but
if
it
does
then
for
some
reason
it
can
be
treated
as
a
separate
thing.
I
don't
think
you
ever
care
about
it
now.
B
A
Nice
yeah,
so
that's
that
makes
things
even
simpler,
then
okay,
and
then
I
guess
this
one
so
yeah
this
part
was
a
little
bit
confused
in
terms
of
the
implementation
itself
and
of
course
I
know,
I
guess
like
thomas
would
know,
know
best,
but
I
was
just
wondering
how,
like
the
different
stages
are
dealt
with,
because
I
know
with
regards
to
the
actual.
A
B
The
question
is:
what
happens
when
you
override
something
in
a
stage,
and
where
does
that
happen
in
the
code.
B
That's
that's
a
good
question,
because
if
you
do
an
ad
resource
using
captain
cli,
it
makes
a
call
to
the
api
and
it
writes
the
resource
for
you
right.
So
if
this
operator
is
also,
I
I
if
it's
doing
if
it's
doing
the
same
thing
and
I
I
feel
like
it's
kind
of
doing
the
same
thing,
so
we
should
not
have
any
override
any
logic
for
our
writing
stuff.
A
B
A
A
B
A
Wouldn't
be
done
via
a
controller,
because
the
controller
is
just
trying
to
like
reconcile
specific
stakes
like
the
actual
states.
Let
me
just
think
about
my
question
yeah.
I
don't
know
why
it
has
this.
B
I
think
I
think
that
might
be,
because
we
have
to
kind
of
we
need
that
this
functionality,
where,
if
there's
some
change
in
personalized
custom
resource,
we
want
to
update
the
resource
in
the
upstream
repository.
A
B
Or
in
the
configuration
service
repository-
and
I
think
that's
where
the
questions
probably
came
from
yeah
captain
overwrites
the
resources.
As
far
as
I
know,
I
could
do
that,
but
this
would
be
a.
B
A
Oh
by
the
way
we
are,
we
are
a
good
overtime
in
case
people
have
to
hop
off.
I
mean
I'm
free
to
stay
a
bit
longer,
but
otherwise
yeah
like
we're.
Definitely
a
bit
over
on
the
30
minutes.
Just
just
an
fyi.
B
B
We
have
a
client.
Let
me,
let
me
think
here
in
slack,
there's
a
thread,
so
we
have
a
client
to
update
the
shipyard.
B
So
you
need
to
figure
out
if
we
can
use
the
same
client,
because
I
kind
of
feel
like
we
should
be
able
to
use
the
same
point.
If
you
look
at
line.
What
is
that
yeah
line
29
could
not
update
status,
ship
yeah,
so
actually
it's
and
if
I
remember
correctly,
I
think
thomas
also
said
we
kind
of
have
that
library
to.
B
List
update
stuff-
and
I
I
can
see
that
we're
using
a
client
for
that
just
need
to
figure
out
if
the
client
also
supports
certain
resources,
which
I
think
it
should
and
if
it
does.
Let
me
just
use
them.
A
Okay,
nice,
okay,
I'll,
definitely
take
a
look
at
that,
and
so
this
will
okay.
So
this
will
actually.
A
Yeah
I
mean
otherwise,
I
think
that's
everything
that
I
wanted
to
to
go
over.
I
mean
thanks
everyone
for
joining
me:
either
live
or
asynchronously
via
video.
A
Yeah,
if
not
thank
you,
everyone
and
I
will
quickly.
A
Oh
thank
you.
Ma'am
I'll
quickly,
conclude
the
recording
here.